The overall goal of this project is to understand how osteocytes, cells embedded in the bone matrix, control the process of bone resorption by osteoclasts under physiological and pathological conditions. Osteoclast formation, function, and survival depend on the cytokine receptor activator of NF?B ligand (RANKL), whose actions are inhibited by the soluble decoy receptor osteoprotegerin (OPG). Recent studies have shown that osteocytes are an important source of the RANKL involved in osteoclast formation. The requirement of osteocyte RANKL for osteoclast formation raises the question of how this cytokine reaches osteoclast progenitors when produced by cells embedded in the bone matrix. RANKL is initially produced as an integral- membrane protein, which can then be cleaved by proteases to yield a soluble form, both of which are able to promote bone resorption. It remains unclear whether the osteoclast formation promoted by osteocyte-derived RANKL involves the membrane-bound form, the soluble form, or both. Cortical porosity is a type of bone loss that increases with aging and glucocorticoid excess, and some studies suggest that the former is associated with increased expression of RANKL in cortical bone. However, it is unknown whether osteocytes are the source of this RANKL and whether this increase is involved in the generation of cortical porosity associated with aging or glucocorticoid excess. Some studies suggest that OPG is also produced by osteocytes and those changes in OPG production may contribute to pathological bone resorption. However, OPG is produced by many cell types and it is unclear which cells provide the OPG important for control of bone remodeling. Based on these observations, we propose the central hypothesis that osteocytes control osteoclast formation via multiple mechanisms including production of soluble RANKL and OPG and this contributes to physiological as well as pathological bone resorption. To address this hypothesis, in Aim 1 we will determine whether osteocytes control osteoclast formation via production of soluble or membrane-bound RANKL. Mice expressing a form of the RANKL protein that is resistant to proteolytic cleavage will be generated and their skeletons analyzed structurally and histologically under physiological and pathological conditions. In Aim 2, we will determine whether osteocyte-derived RANKL contributes to the cortical porosity caused by aging or glucocorticoid excess. Bone mass and structure, including cortical porosity, as well as osteoclast number, will be measured in 20-month-old mice lacking the RANKL gene in osteocytes and control littermates. The impact of glucocorticoid excess on cortical porosity will also be examined in adult mice of the same genotypes. Lastly, in Aim 3 we will determine whether osteocytes are an important source of OPG. A conditional allele of the OPG gene will be deleted from osteocytes using Dmp1-Cre and Sost-Cre transgenic mice and the skeletal phenotype compared to that of mice with germline deletion of OPG.